PTEN Alterations Are Related to Unfavorable Prognosis in Children with T-Cell Acute Lymphoblastic Leukemia Treated According to ALL IC-BFM Protocols

Background: Resistant or relapsed T-cell acute lymphoblastic leukemia (T-ALL) has a poor prognosis and is usually fatal. Thus, patients with a very high risk of treatment failure need to be considered for alternative treatments. To identify these patients, clinical parameters and response to therapy are currently in use as prognostic markers. These markers, however, do not allow for reliable identification of all high-risk patients early during induction treatment. In search for genetic markers that inform treatment strategies early after leukemia diagnosis, we evaluated the prognostic significance of a panel of 25 established T-ALL oncogenes or tumor suppressor genes, in children treated according to ALL IC-BFM protocols.

Patients and Methods: The study group included 162 children with de novo T-ALL, treated according to I-BFM protocols ALL-IC BFM 2002 (n=91; 56%) and ALL-IC BFM 2009 (n=71; 44%). Mutations in NOTCH1 (exons 26, 27, 34) , FBXW7 (exons 9 and 10) , PTEN (exon 5, 6, 7) , WT1 (exon 7 and 9) , IL7R (exon 6) , STAT5B (exon 16) , FLT3 (exons 11-12) , RUNX1 (exons 1, 2, 3, 6) and DNMT3A (exons 8-23) were detected by either high resolution melt analysis (HRM), Sanger sequencing, or both. Copy number alterations (CNAs) in SIL-TAL1, LEF1, CASP8AP2, MYB, EZH2, CDKN2A/B, MLLT3, NUP214-ABL1, LMO1, LMO2, NF1, SUZ12, PTPN2, PTEN, PHF6 were detected by multiplex ligation-dependent probe amplification (MLPA) using SALSA MLPA P383 T-ALL probe mix (MRC Holland) and a custom made probes for exons 2, 3, 4, 5 of PTEN . MRD was assessed by multicolor flow cytometry (FC-MRD at d15, 33, 78) and 'FC-MRD positive' result was defined as >=0.01%. Statistical analyses: Fisher exact test, Cochran-Mantel-Haenszel test (CMH), Kaplan-Meier estimators and log-rank test. Probability of relapse-free survival (pRFS) was calculated using the time from diagnosis to the date of relapse or the last follow-up in complete remission.

Results: Of all 25 genes analyzed, PTEN alterations significantly predict for the risk of relapse in 162 children treated according to ALL IC-BFM protocols. The strength of the study is combined analysis of CNAs and point mutations that enabled us to identify PTEN alterations (PTEN.ALT), resulting from both loss-of function mutations (PTEN.MUT) and/or genomic deletions (PTEN.DEL). Patients with PTEN.ALT were at higher risk of relapse with pRFS of 60% vs. 83% in non-mutated patients, p=0.0053. Moreover,in the separate analysis of each of PTEN alterations (PTEN.DEL or PTEN.MUT), both were associated with worse pRFS (58% in patients with PTEN.DEL vs. 82% in non-mutated, p=0.01; and pRFS of 60% in patients with PTEN.MUT vs. 82% in non-mutated, p=0.06). However, in the analysis of PTEN status (PTEN.ALT, PTEN.MUT or PTEN.DEL) in each of the treatment protocols separately, the significant impact on outcome was retained only in the ALL IC-BFM 2009 protocol. Remarkably, long-term negative impact of PTEN alterations was reflected in the poorer early response to treatment. The odds ratios for the occurrence of unfavorable prednisone response (≥ 1000 blasts/mL) were: 3.06 (95%CI: 1.22-7.67, p=0.02) among patients with PTEN.ALT; 3.49 (95%CI: 1,21-10,03, p=0.02) among patients with PTEN.DEL and 3.99 (95%CI: 1,13-13,36, p=0.03) among patients with PTEN.MUT. In addition, in a stratified analysis for both ALL-IC protocols, we found patients with PTEN.MUT to be associated with positive FC-MRD at day 78 (OR: 11,13, 95%CI: 2.38-51.92, p=0.0033). With respect to characteristics at diagnosis,patients with PTEN.DEL or PTEN.ALT status more often presented with higher WBC counts (>=20 000) than patients with wild-type PTEN (p=0.01 and p=0.02, respectively) and those with PTEN.MUT status were predominantly males (P=0.03).

Conclusion: Out of the 25 established T-ALL driver genes analyzed in 162 children treated according to ALL IC-BFM schemes, PTEN abnormalities predict worse relapse-free survival and poorer early response to treatment and are candidates for genetic risk markers in pediatric T-ALL.

Research funded by National Science Centre, Poland grant 2013/11/N/NZ5/03730 and National Centre of Research and Development (NCRD) grant STARTEGMED3/304586/5/NCBR/2017.